This project aims to identify alternately spliced transcripts from the DNA-repair genes from lung and breast tissue that perturb the endogenous DNA-repair mechanism. A number of alternately spliced transcripts of the DNA-repair genes have been identified from adult mouse and human tissue. Interestingly, these variants disrupt DNA repair in a model system, even in the presence of the wild-type protein. To extend this study, an mRNA splicing screen of the genes responsible for error free DNA repair in mammalian cells will be undertaken. Both normal and tumor (adenocarcinoma) tissue from clinical samples will be screened using a splice-specific RT-PCR protocol. Using this procedure all of the alternately spliced DNA-repair transcripts from these tissues can be characterized. Typically mutations in the DNA-repair genes are lethal to the cell. However, exciting preliminary data shows that alternatively spliced transcripts can encode novel proteins, which modify the DNA-repair activity of the cell. To identify splice variants that disrupt DNA repair, a rapid functional screen using a yeast model system will be performed. Yeast is used as the DNA-repair paradigm for eukaryotes and its DNA-repair activities are functionally conserved with mammals. Select variants that dominantly disrupt DNA-repair in yeast will then be tested in mammalian cell culture to ensure that the effects seen in yeast are conserved in human cells. In this way, splice variants of the DNA-repair genes that dominantly disrupt DNA repair and potentially lead to tumorigenesis will be identified. The alternately spliced transcripts identified in this screen will provide candidate biomarkers for tumorigenesis. Based upon the data from this study, it is anticipated that a larger screen of clinical samples will be conducted to establish a direct association with a specific tumor type. Such a screen is beyond the scope of this R21 application. By focusing upon splicing, a new class of biological regulation may be uncovered that profoundly affects gene function, but is normally overlooked in other clinical screens. In addition, the approach outlined here is applicable to other tissue types and it is anticipated that this research will lead to further study of other human tumors.